TMDL Summary Appendix 666---111777

Chandler River Watershed Description Waterbody Facts This TMDL applies to a 27.19 mile section of the Chandler River, Segment ID: including its eastern branch, located in the Towns of Duram, ME0106000102_603R02 Pownal and North Yarmouth, . The headwaters of the Town: Duram, Pownal, Chandler River are at the outlet of Runaround Pond in Durham. North Yarmouth, ME The river flows south, crossing Runaround Pond Road, then County: Cumberland through a wooded area before crossing Poland Range Road, Lawrence Road, Elmwood Road, and Chadsey Road, in an area of Impaired Segment mixed forest, agriculture and residential development. The river Length: 27.19 miles continues south, crossing Milliken Road, and meets with the East Classification: Class B Branch before flowing south to its convergence with the in Gray. The East Branch originates in a wetland complex in Direct Watershed: 51.89 2 Durham. The river flows south through a predominantly forested mi (33,210 acres) area before crossing Quaker Meeting House Road, Brown Road, Impairment Listing Poland Range Road, Tuttle Road and Elmwood Road. The river Cause: Dissolved Oxygen then crosses Hodsdon Road, flows north across Hallowell Road and then south again. Downstream of the West Pownal Road crossing, Watershed Agricultural the East Branch meets the main stem of the Chandler River. Land Use: 14.3% The Chandler River watershed covers an area of 51.89 square Major Drainage Basin: miles. The majority of the watershed is located with0in the Town of Pownal, however, smaller portions of the watershed lie within the surrounding towns of Auburn, New Gloucester, Durham, Brunswick, Freeport, Gray, and North Yarmouth. ‹ Runoff from agricultural land along Tuttle Road is likely the largest source of nonpoint source (NPS) pollution to the Chandler River. Runoff from cultivated lands, active hay lands, and pasture can transport nitrogen and phosphorus to the nearest section of the stream. ‹ The Chandler River watershed is predominately non-developed (95.8%). Forested areas (74.1%) within the watershed absorb and filter pollutants helping protect both water quality in the stream and stream channel stability. Wetlands (7%) may also help filter nutrients. ‹ Non-forested areas within the watershed are predominantly agricultural (14.3%). Developed areas (4.2%) with impervious

surfaces may impact water quality.

‹ The Chandler River including its East Branch is on Maine’s 303(d) Watershed Land Uses list of Impaired Streams (Maine DEP, 2013).

Definitions Agriculture • Total Maximum Daily Load (TMDL) represents the total amount Forest of pollutants that a waterbody can receive and still meet water Wetland quality standards. Developed • Nonpoint Source Pollution refers to pollution that comes from many diffuse sources across the landscape, and is typically transported by rain or snowmelt runoff. Maine Statewide TMDL for NPS Pollution June 2016

Figure 1: Land Use in the Chandler River Watershed

WHY IS A TMDL ASSESSMENT NEEDED ?

The Chandler River including its East Branch, a Class B freshwater stream, has been assessed by Maine DEP as not meeting water quality standards for the designated use of aquatic life, and placed on the 303(d) list of impaired waters under the Clean Water Act. The Clean Water Act requires that all 303(d)-listed waters undergo a TMDL assessment that describes the impairments and establishes a target to guide the measures needed to restore water quality. The goal is for all waterbodies to comply with state water quality standards.

Agriculture in the Chandler River watershed makes up about 14% of the total land area, more than three times the area of Chandler River (East Branch) at Tuttle Road. Photo: FB Environmental developed land which accounts for about 4% of total land area (Figure 1). Agriculture is therefore likely to be the largest contributor of sediment and nutrient enrichment to the stream. The close proximity of many agricultural lands to the stream further increases

APPENDIX 6-17 2 Maine Statewide TMDL for NPS Pollution June 2016 the likelihood that nutrients from disturbed soils, manure, and fertilizers will reach the stream. Along Tuttle Road, in particular, heavy erosion was documented as a result of livestock in the stream.

WATER QUALITY DATA ANALYSIS

Maine DEP uses a variety of data types to measure the ability of a stream to adequately support aquatic life, including; dissolved oxygen, benthic macroinvertebrates, and periphyton (algae). The aquatic life impairment in Chandler River is based on historic dissolved oxygen data.

TMDL ASSESSMENT APPROACH : NUTRIENT MODELING OF IMPAIRED AND ATTAINMENT STREAMS

NPS pollution is difficult to measure directly, because it comes from many diffuse sources spread across the landscape. For this reason, a nutrient loading model, MapShed, was used to estimate the sources of pollution based on well-established hydrological equations; detailed maps of soil, land use, and slope; many years of daily weather data; and direct observations of agriculture and other land uses within the watershed.

The nutrient loading estimates for the impaired stream were compared to similar estimates for five non- impaired (attainment) streams of similar watershed land uses across the state. The TMDL for the impaired stream was set as the mean nutrient loading estimate of these attainment stream watersheds, and units of mass per unit watershed area per year (kg/ha/year) were used. The difference in loading estimates between the impaired and attainment watersheds represents the percent reduction in nutrient loading required under this TMDL. The attainment streams and their nutrient and sediment loading estimates and TMDL are presented below in Table 1.

Table 1: Numeric Targets for Pollutant Loading Based on MapShed Model Outputs for Attainment Streams TP load TN load Sediment load Attainment Streams Town (kg/ha/yr) (kg/ha/yr) (1000 kg/ha/yr) Martin Stream Fairfield 0.14 3.4 0.008 Footman Brook Exeter 0.33 6.4 0.058 Upper Corinth 0.29 5.6 0.047 Upper Pleasant River Gray 0.22 4.6 0.016 Moose Brook Houlton 0.25 5.9 0.022 Total Maximum Daily Load 0.24 5.2 0.030

APPENDIX 6-17 3 Maine Statewide TMDL for NPS Pollution June 2016

RAPID WATERSHED ASSESSMENT

Habitat Assessment A Habitat Assessment survey was conducted on both the impaired and attainment streams. The assessment approach is based on the Rapid Bioassessment Protocols for Use in Wadeable Streams and Rivers (Barbour et al., 1999), which integrates various parameters relating to the structure of physical habitat. The habitat assessments include a general description of the site and physical characterization and visual assessment of in-stream and riparian habitat quality.

Based on Rapid Bioassessment protocols for low gradient streams, the Chandler River received a score of 171 out of a total 200 for quality of habitat. Higher scores indicate better habitat. The range of habitat assessment scores for attainment streams was 155 to 179. RAPID HABITAT ASESSMENT SCORES The habitat assessment was conducted on a for Attainment and Impaired Streams relatively short sample reach (about 100-200 meters 200 for a typical small stream), and was located near the most downstream Maine DEP sample station. For 190 both impaired and attainment streams, the assessment location was usually near a road 180 crossing for ease of access. In the Chandler River watershed, the downstream sample station was 170 located in a forested portion of the stream with a thick buffer. Although there is some agriculture 160 Attainment within the Chandler River watershed, a majority of Impaired the stream flows within forested areas. 150 Figure 2 (right) shows the range of habitat Habitat Score Chandler 140 assessment scores for all attainment and impaired River streams, as well as for the Chandler River. The overlapping attainment and impaired stream scores 130 indicate that factors other than habitat should be considered when addressing the impairments in the 120 Chandler River. Consideration should be given to major “hot spots” in the Chandler River watershed 110 as potential sources of NPS pollution contributing to the water quality impairment. 100

Figure 2: Habitat Assessment Scores

Pollution Source Identification Pollution source identification assessments were conducted for both Chandler River (impaired) and the attainment streams. The source identification component of this study is based on an abbreviated version of the Center for Watershed Protection’s Unified Subwatershed and Site Reconnaissance method (Wright, et al., 2005). The abbreviated method includes both a desktop and field component. The desktop assessment consists of generating and reviewing maps of the watershed boundary, roads, land use and satellite imagery, and then identifying potential NPS pollution locations, such as road crossings, agricultural fields, and large areas of bare soil. When available, multiple sources of satellite imagery

APPENDIX 6-17 4 Maine Statewide TMDL for NPS Pollution June 2016

were reviewed. Occasionally, the high resolution of the imagery allowed for observations of livestock, row crops, eroding stream banks, sediment laden water, junkyards, and other potential NPS concerns that could affect stream quality. As many potential pollution sources as possible were visited, assessed and documented in the field. Field visits were limited to NPS sites that were visible from roads or a short walk from a roadway. Neighborhoods were assessed for NPS pollution at the whole neighborhood level including streets and storm drains (where applicable). The assessment does not include a scoring component, but does include a detailed summary of findings and a map indicating documented NPS sites throughout the watershed.

The watershed source assessment for the Chandler River was completed on July 6, 2012. In-field observations of erosion, lack of vegetated stream buffer, extensive impervious surfaces, high-density neighborhoods and agricultural activities were documented throughout the watershed (Table 1, Figure 3).

Table 2: Pollution Source ID Assessment for the Chandler River Watershed

Potential Source ID# Location Type Notes Road • Moderate erosion observed at road crossing. Lawrence 9 Crossing/ • Road Residential property near crossing has a small pasture with Agriculture approximately 5 sheep. • Severe erosion was observed at the Chadsey Road crossing. • The portion of roadway over bridge is paved though all roads leading to bridge are gravel/dirt roadways. Chadsey Road • Some sediment is captured by a grassy buffer near the stream and a 14 Road Crossing ditch leading to the crossing. • It is difficult to determine if all sediment is from unpaved roads or construction activity occurring nearby at a power line maintenance staging area. • Severe sedimentation due to livestock entering river. Agriculture • No livestock was observed during field visit, however, ample 26 & Tuttle /Road evidence supports observations, e.g. trodden paths with direct 27 Road Crossing access to the river, hoof prints on stream substrates, and large slumping banks contributing to sediment build-up within the river. • Moderate road shoulder erosion along Elmwood Road is causing sedimentation at the road crossing. Elmwood • Unstable areas were marked by DPW sawhorses at some locations. Road • Road A steep slope lies between Elmwood Road and Chandler River on 39 (east of Crossing the west (downstream) side of the road. Lawrence • A field enclosed by a livestock fence was observed near the Road) crossing. This field lies within close proximity to the river’s edge with a very small vegetated buffer. No livestock was observed during the field visit.

APPENDIX 6-17 5 Maine Statewide TMDL for NPS Pollution June 2016

Figure 3: Aerial Photo of Source ID Locations in the Chandler River Watershed

APPENDIX 6-17 6 Maine Statewide TMDL for NPS Pollution June 2016

NUTRIENT LOADING – MAP SHED ANALYSIS The MapShed model was used to estimate stream loading of sediment, total nitrogen and total phosphorus in the Chandler River (impaired) plus five attainment watersheds throughout the state. The model estimated nutrient loads over a 15-year period (1990-2004), which was determined by the available weather data provided within MapShed. This extended period captures a wide range of hydrologic conditions to account for variations in nutrient and sediment loading over time.

Many quality assured and regionally calibrated input parameters are provided with MapShed. Additional input parameters were manually entered into the model based on desktop research and field observations, as described in the sections on Habitat Assessment and Pollution Source Identification. These manually adjusted parameters included estimates of livestock animal units, agricultural stream miles with intact vegetative buffer, Best Management Practices (BMPs), and estimated wetland retention and/or drainage areas.

Livestock Estimates Livestock waste contains nutrients which can cause water quality Table 3 : Livestock Estimates in impairment. The nutrient loading model considers numbers and the Chandler River Watershed types of animals. Table 3 (right) provides estimates of livestock Chandler Type (numbers of animals) in the watershed, based on direct observations River made in the watershed, plus other publicly available data. Dairy Cows 30 Beef Cows

There are agricultural areas throughout the Chandler River Broilers

watershed. Several NPS pollution sites were documented in the field Layers

including livestock. A cow farm located on Tuttle Road adjacent to Hogs/Swine 20 the Tuttle Road stream crossing is a hotspot of pollution entering the Sheep 5 Chandler River. At this location stream banks are significantly Horses 18 slumping and eroded due to livestock accessing the stream. The Turkeys river flows through unfenced pasture, hoof prints and manure were Other observed in the river and on the banks. Significant sedimentation was observed downstream of the crossing. A strong manure smell Total 73 was also present.

A pig breeding farm is located on Knoll Hill Farm on Hallowell Road. No livestock were observed, but 20 pigs were estimated based on multiple signs reading “pigs and piglets for sale.” Eighteen horses were observed throughout the watershed on two properties on West Pownal Road and Grant Road. Five sheep were also observed grazing along Lawrence Road.

APPENDIX 6-17 7 Maine Statewide TMDL for NPS Pollution June 2016

Vegetated Stream Buffer in Agricultural Areas Vegetated stream buffers are areas of trees, shrubs, and/or grasses Table 4: Summary of Vegetated adjacent to streams, lakes, ponds or wetlands which provide Buffers in Agricultural Areas nutrient loading attenuation (Evans & Corradini, 2012). MapShed Chandler River considers natural vegetated stream buffers within agricultural areas • 82.8 stream miles in as providing nutrient load attenuation. The width of buffer strips is watershed (includes ephemeral not defined within the MapShed manual, and was considered to be streams) 75 feet for this analysis. Geographic Information System (GIS) analysis of recent aerial photos along with field reconnaissance • 1.74 stream miles in observations were used to estimate the number of agricultural agricultural areas stream miles with and without vegetative buffers, and these • 98% of agricultural stream estimates were directly entered into the model. miles have a vegetated buffer The Chandler River including its East Branch is a 27.2 mile-long impaired segment as listed by Maine DEP. As modeled, the total stream miles (including tributaries) within the watershed was calculated as 82.8 miles. Of this total, 1.74 stream miles are located within agricultural areas; of this length, 1.71 miles (98%) show a 75-foot or greater vegetated buffer (Table 4, Fig. 4). By contrast, agricultural stream miles (as modeled) with a 75-foot vegetated buffer in the attainment stream watersheds ranged from 34% to 92%, with an average of 61%.

APPENDIX 6-17 8 Maine Statewide TMDL for NPS Pollution June 2016

Figure 4: Buffered Agricultural Stream Miles in the Chandler River Watershed

APPENDIX 6-17 9 Maine Statewide TMDL for NPS Pollution June 2016

Best Management Practices (BMPs) For this modeling effort, four commonly used BMPs were entered based on literature values. These estimates were applied equally to impaired and attainment stream watersheds. More localized data on agricultural practices would improve this component of the model.

• Cover Crops: Cover crops are the use of annual or perennial crops to protect soil from erosion during time periods between harvesting and planting of the primary crop. The percent of agricultural acres cover crops used within the model is estimated at 4%. This figure is based on information from the 2007 USDA Census stating that 4.1% of cropland acres is left idle or used for cover crops or soil improvement activity, and not pastured or grazed (USDA, 2007b). • Conservation Tillage: Conservation tillage is any kind of system that leaves at least 30% of the soil surface covered with crop residue after planting. This reduces soil erosion and runoff and is one of the most commonly used BMPs. This BMP was assumed to occur in 42% of agricultural land. This figure is based on a number given by the Conservation Tillage Information Center’s 2008 Crop Residue Management Survey stating that 41.5% of U.S. acres are currently in conservation tillage (CTIC, 2000). • Strip Cropping / Contour Farming: This BMP involves tilling, planting and harvesting perpendicular to the gradient of a hill or slope using high levels of plant residue to reduce soil erosion from runoff. This BMP was assumed to occur in 38% of agricultural lands, based on a study done at the University of Maryland (Lichtenberg, 1996). • Grazing Land Management: This BMP consists of ensuring adequate vegetation cover on grazed lands to prevent soil erosion from overgrazing or other forms of over-use. This usually employs a rotational grazing system where hays or legumes are planted for feed and livestock is rotated through several fenced pastures. In this TMDL, a figure of 75% of hay and pasture land is assumed to utilize grazing land management. This figure is based on a study by Farm Environmental Management Systems of farming operations in Canada (Rothwell, 2005).

Pollutant Load Attenuation by Lakes, Ponds and Wetlands Depositional environments such as ponds and wetlands can attenuate watershed sediment loading. This information is entered into the nutrient loading model by a simple percentage of watershed area draining to a pond or a wetland. The Chandler River watershed is 7% wetland, and overall 5% of the watershed drains to wetlands. Percent of watershed draining to a wetland in the attainment watersheds ranged from 15% to 60%, with an average of 35%.

NUTRIENT MODELING RESULTS

The MapShed model simulates surface runoff using daily weather inputs of rainfall and temperature. Erosion and sediment yields are estimated using monthly erosion calculations and land use/soil composition values for each source area. Below, selected results from the watershed loading model are presented. The TMDL itself is expressed in units of kilograms per hectare per year. The additional results shown below assist in better understanding the likely sources of pollution. The model results for the Chandler River indicate that significant reductions of sediment and nutrients are needed to improve water quality. Below, loading for sediment, nitrogen and phosphorus are discussed individually.

APPENDIX 6-17 10 Maine Statewide TMDL for NPS Pollution June 2016

Sediment Table 5: Total Sediment Loads by Source Sediment Sediment Sediment loading in the Chandler Chandler River River watershed is mainly derived (1000kg/year) (%) from hay/pasture which contributes Source Load 37% of the total load. Forested lands Hay/Pasture 145.32 37% are also a main source of sediment Crop land 53.93 14% accounting for 35% of the total Forest 135.30 35% sediment load. Total loads by mass Wetland 3.08 1% cannot be directly compared between Disturbed Land 4.32 1% watersheds due to differences in Low Density Mixed 9.46 2% watershed area. See section TMDL: Medium Density Mixed 0 0% Target Nutrient Levels for the High Density Mixed 39.38 10% Chandler River (below) for loading Low Density Residential 1.25 0% estimates that have been normalized Medium Density Residential 0 0% by watershed area. High Density Residential 0 0% Farm Animals 0 0% Septic Systems 0 0% Source Load Total: 392.04 100%

Pathway Load

Stream Banks 331.13 - Subsurface / Groundwater 0.00 -

Total Watershed Mass Load: 723.17

Sediment Load by Source 40%

30%

20%

10% Total Sediment 0%

Sediment Sources

Figure 5: Total Sediment Loads by Source in the Chandler River Watershed

APPENDIX 6-17 11 Maine Statewide TMDL for NPS Pollution June 2016

Total Nitrogen Table 6: Total Nitrogen Loads by Source Total N Total N Nitrogen loading in the Chandler River Chandler River watershed is mainly attributed to (kg/year) (%) forested lands which account for 41% Source Load Hay/Pasture 2895.4 19% of the total load. Hay/pasture and septic Crop land 1024.6 7% systems also contribute a significant Forest 6419.5 41% percent of the load with 19% and 13%, Wetland 883.5 6% respectively. Table 6 and Figure 6 Disturbed Land 18.3 0% (below) show the estimated total Low Density Mixed 251.1 2% nitrogen load in terms of mass and Medium Density Mixed 0 0% percent of total by source in the High Density Mixed 1584.2 10% Chandler River. Total loads by mass Low Density Residential 33.2 0% cannot be directly compared between Medium Density Residential 0 0% watersheds due to differences in High Density Residential 0 0% watershed area. See section TMDL: Farm Animals 503.9 3% Target Nutrient Levels for the Chandler Septic Systems 2010.7 13% River (below) for loading estimates that Source Load Total: 15624.5 100% have been normalized by watershed area. Pathway Load Stream Banks 177.9 - Subsurface / Groundwater 62525.8 -

Total Watershed Mass Load: 78328.2

TN Load by Source 50% 40% 30% 20% Total N 10% 0%

TN Sources

Figure 6: Total Nitrogen Loads by Source in the Chandler River Watershed

APPENDIX 6-17 12 Maine Statewide TMDL for NPS Pollution June 2016

Total Phosphorus Table 7: Total Phosphorus Loads by Source Total P Total P Phosphorus loading in the Chandler Chandler River River watershed is attributed primarily (kg/year) (%) to hay/pasture lands (51%). Source Load Phosphorus loads are presented in Hay/Pasture 1101.8 51% Table 7 and Figure 7. Total loads by Crop land 135.6 6% mass cannot be directly compared Forest 424.2 19% between watersheds due to differences Wetland 48.1 2% in watershed area. See section TMDL: Disturbed Land 8.0 0% Target Nutrient Levels for the Low Density Mixed 28.3 1% Chandler River (below) for loading Medium Density Mixed 0 0% estimates that have been normalized High Density Mixed 164.9 8% by watershed area. Low Density Residential 3.7 0% Medium Density Residential 0 0% High Density Residential 0 0% Farm Animals 99.1 5% Septic Systems 165.6 8% Source Load Total: 2179.1 100%

Pathway Load Stream Banks 65.0 - Subsurface / Groundwater 1631.2 -

Total Watershed Mass Load: 3875.4

TP load by Source 60% 50% 40% 30%

Total P 20% 10% 0%

Sources of TP

Figure 7: Total Phosphorus Loads by Source in the Chandler River Watershed

APPENDIX 6-17 13 Maine Statewide TMDL for NPS Pollution June 2016

TMDL: TARGET NUTRIENT LEVELS FOR THE CHANDLER RIVER The existing sediment and nutrient loads for the impaired segment of the Chandler River are listed in Table 8, along with the TMDL numeric target which was calculated from the average loading estimates of five attainment watersheds throughout the state. Table 9 presents a more detailed view of the modeling results and calculations used in Table 8 to define TMDL reductions, and compares the existing sediment and nutrient loads in the Chandler River to TMDL endpoints derived from the attainment waterbodies. An annual time frame provides a mechanism to address the daily and seasonal variability associated with nonpoint source loads.

Table 8: TMDL Targets Compared to the Chandler River Pollutant Loading

Estimated Loads TMDL % TMDL POLLUTANT LOADS Total Maximum Daily for REDUCTIONS Annual Loads per Unit Area Load Numeric Target Chandler River Chandler River Sediment Load (1000 kg/ha/year) 0.054 0.030 45% Nitrogen Load (kg/ha/year) 5.88 5.2 12% Phosphorus Load (kg/ha/year) 0.29 0.24 16%

Future Loading The prescribed reduction in pollutants discussed in this TMDL reflects reduction from estimated existing conditions. Expansion of agricultural and development activities have the potential to increase runoff and associated pollutant loads to the Chandler River. To ensure that the TMDL targets are attained, future agriculture or development activities in the watershed will need to meet the TMDL targets. Future growth from population increases is a moderate threat in the Chandler River watershed because Cumberland County has increasing population trends, with a 3.9% increase between 2000 and 2008 (USM MSAC, 2009). The growth in agricultural lands is also increasing, with a 6% increase in the total number of farms in Cumberland County between 2002 and 2007. However, a decrease of 5% was seen in the land (acres) in farms between 2002 and 2008, and a 10% decrease occurred in the average farm size in this time period as well (USDA, 2007a). Future activities and BMPs that achieve TMDL reductions are addressed below.

Next Steps The use of agricultural and developed area BMPs can reduce sources of polluted runoff in the Chandler River. It is recommended that municipal officials, landowners, and conservation stakeholders in Duram, Pownal and North Yarmouth work together to develop a watershed management plan to:

‹ Encourage greater citizen involvement through the development of a watershed coalition to ensure the long term protection of Chandler River; ‹ Address existing nonpoint source problems in the Chandler River watershed by instituting BMPs where necessary; and ‹ Prevent future degradation of the Chandler River through the development and/or strengthening of a local Nutrient Management Ordinance.

APPENDIX 6-17 14

Maine Statewide TMDL for NPS Pollution June 2016

Table 9: Modeling Results Calculations for Derived Numeric Targets and Reduction Loads for the Chandler River Chandler River Area Sediment TN TP ha 1000kg/yr kg/yr kg/yr Land Uses Hay/Pasture 1765 145.3 2895.4 1101.8 Crop land 147 53.9 1024.6 135.6 Forest 9805 135.3 6419.5 424.2 Wetland 933 3.1 883.5 48.1 Disturbed Land 73 4.3 18.3 8.0 Low Density Mixed 257 9.5 251.1 28.3 High Density Mixed 311 39.4 1584.2 164.9 Low Density Residential 34 1.3 33.2 3.7 Other Sources

Farm Animals 503.9 99.1

Septic Systems 2010.7 165.6

Pathway Loads

Stream Banks 331.1 177.9 65.0

Groundwater 62525.8 1631.2

Total Annual Load 723 x 1000 kg 78328 kg 3875 kg

Total Area 13326 ha

Total Maximum Daily 0.054 5.88 0.29

Load 1000kg/ha/year kg/ha/year kg/ha/year

APPENDIX 6-17 15 Maine Statewide TMDL for NPS Pollution June 2016

REFERENCES Barbour, M.T., J. Gerritsen, B.D. Snyder, and J.B. Stribling. 1999. Rapid Bioassessment Protocols for Use in Wadeable Streams and Rivers: Periphyton, Benthic Macroinvertebrates and Fish, Second Edition. EPA 841-B-99-002. U.S. Environmental Protection Agency; Office of Water; Washington, D.C. Conservation Tillage Information Center (CTIC). 2000. Crop Residue Management Survey. National Association of Conservation Districts. Retrieved from: http://www.ctic.purdue.edu . Davies, S. P., and L. Tsomides. 2002. Methods for Biological Sampling of Maine's Rivers and Streams. DEP LW0387-B2002, Maine Department of Environmental Protection, Augusta, ME. Evans, B.M., & K.J. Corradini. 2012. MapShed Version 1.0 Users Guide. Penn State Institute of Energy and the Environment. Retrieved from: http://www.mapshed.psu.edu/Downloads/MapShedManual.pdf Lichtenberg, E. 1996. Using Soil and Water Conservation Practices to Reduce Bay Nutrients: How has Agriculture Done? Economic Viewpoints. Maryland Cooperative Extension Service, University of Maryland at College Park and University of Maryland Eastern Shore, Department of Agricultural and Resource Economics, 1(2). Maine Department of Environmental Protection (Maine DEP). 2013. Draft 2012 Integrated Water Quality Monitoring and Assessment Report. Bureau of Land and Water Quality, Augusta, ME. Rothwell, N. 2005. Grazing Management in Canada. Farm Environmental Management in Canada. http://publications.gc.ca/Collection/Statcan/21-021-M/21-021-MIE2005001.pdf . University of Southern Maine Muskie School of Public Service, Maine Statistical Analysis Center (USM MSAC). December, 2009. Retrieved from: http://muskie.usm.maine.edu/justiceresearch/Publications/County/Cumberland.pdf United States Department of Agriculture (USDA). 2007a. 2007 Census of Agriculture: Cumberland County, Maine. Retrieved from: http://www.agcensus.usda.gov/Publications/2007/Online_Highlights/County_Profiles/Maine/cp2 3005.pdf United States Department of Agriculture (USDA). 2007b. 2007 Census of Agriculture: State and County Reports. National Agricultural Statistics Service. Retrieved from: http://www.agcensus.usda.gov/Publications/2007/Full_Report/Volume_1,_Chapter_1_State_Lev el/Maine/st23_1_008_008.pdf Wright, T., C. Swann, K. Cappiella, and T. Schueler. 2005. Unified Subwatershed and Site Reconnaissance: A User’s Manual. Center for Watershed Protection. Ellicott City, MD.

APPENDIX 6-17 16